A requirement for conventional binary data transmission is the establishment of a well-defined logic threshold. Metallic wired systems use pre-defined dc logic levels for this purpose. This is unsatisfactory in an optical system where absolute signal levels are commonly not known a priori. The conventional solution is ac-coupling between the receiver and the logic quantizer. With this approach, dc logic threshold levels are established by forming a “signal average” of the received data pulse. Signals above the average are considered as logic ONEs, while signals below the average are logic ZEROs. While ac-coupled receivers may work well for encoded continuous data transmission they do not work well for burst mode data transmissions and signals where the time average of the signal may vary.
A well known problem in the art of digital optical communication is the difficulty in using unencoded data, i.e. a direct unencoded transmission of data. Here the data to be transmitted is allowed to have long strings of only ONEs or only ZEROs also referred to as long CIDs (Consecutive Identical Digits) and/or the data is not necessarily DC balanced (i.e. having on average substantially equal amounts of the high logic and low logic values). This is because, for proper operation of an AC coupled receiver, the optimal threshold is usually substantially midway between the logic high and logic low value and should substantially correspond to the average (for brevity “logic high” and “logic low” may be referred to as simply high and low values). Therefore non-transition periods, i.e. a period of long CIDs should be small relative to the time constant determined by the capacitance for coupling between a) a preamplifier of an electrically-converted version of the optical signal and b) a comparator which is used to determine the logic value, i.e., 1 or 0, of the incoming optical signal. In optical links this problem is commonly solved by using encoding of the signal prior to transmission and decoding after reception. The encoding ensures that transition occurs in the optical data signal even if the data is a long string of CIDs. Furthermore, the encoding often ensures that the optical signal is balanced. Disadvantageously, doing so mandates that a decoder be present in the receiving system to remove the formatting and reconstruct the original data. Also, data transmission efficiency is commonly degraded because of the required extra bits for the encoding.
While AC coupling is not always needed in electrical communication systems, it was believed to be required in optical systems, because there is no common electrical connection between the transmitter and the receiver. Nevertheless, eventually some of the above-noted disadvantages were overcome by burst-mode digital optical receivers which use direct current (DC) coupling. Such a DC coupled burst-mode digital optical receivers are disclosed in U.S. Pat. No. 5,025,456 which was intended to adapt to the amplitude of the incoming burst data packet and automatically adjust the logic threshold voltage to the dc center, ideally during the first bit of the input data burst. The inventors later characterized the purpose of this invention as meeting the challenge of having several transmitters on the same optical bus where the power levels may vary dramatically between transmitters, see U.S. Pat. No. 5,430,766. This later patent aimed at solving the problem of having a relatively high amount of constant light from each transmitter combined reaching the receiver. The same inventors also patented a burst mode receiver in U.S. Pat. No. 5,371,763. The above cited patents are incorporated herein by reference.
However, one drawback of these prior arts burst-mode receivers is that they rely on the use of a preamble preceding the payload data in a data burst to set the logic threshold. Such a preamble will often comprise a data sequence aimed at initializing the receiver rather than carry data. Such receivers are therefore often not suitable for receiving unencoded data because correct determination of the first digits after a long CID may not be ensured at least not with good pulse width distortion.